Perforce Delphix launches AI-native synthetic data platform
Perforce has launched Delphix Synthetic Data, an AI-native platform designed to generate scenario-specific test data that preserves referential integrity across enterprise databases and file formats. The product targets development and testing teams blocked by restricted, incomplete, or non-existent production data, as well as agentic workflows where AI agents need safe, high-quality data environments to operate.
The platform uses AI to scan metadata and schemas automatically, applying statistical analysis to understand data shape and distribution before generating synthetic datasets. Users can specify and adjust the data they need through natural language prompts, cutting what the company describes as days or weeks of waiting for low-quality test data from legacy tools down to minutes. A bring-your-own-LLM option means generated data never leaves an organisation's own infrastructure, a feature designed to address data-sovereignty and compliance requirements.
What the survey found
Perforce commissioned a survey of 518 enterprise technology leaders in 2026 to support the launch. Among respondents who had evaluated existing synthetic data solutions, only 34% said those tools delivered referential integrity, and only 36% said they provided adequate data realism. Both capabilities are, by the company's own framing, essential for production-grade development, testing, and agentic pipelines.
Jim Mercer, Programme Vice President for Software Development, DevOps, and DevSecOps at IDC, commented that organisations need test data that can be generated quickly, scaled across complex environments, and meet privacy requirements. He positioned solutions combining AI-driven automation with data quality controls as better placed to address that need.
Ilker Taskaya, Field CTO at Perforce Delphix, put the business case plainly: masked production data only reflects historical states, whereas testing increasingly requires edge cases that do not yet exist in production, delivered with intact relationships across every database and file format in the environment.
Delphix Synthetic Data is integrated into the broader Delphix DevOps Data Platform, which bundles data masking, data delivery, and centralised governance alongside the new synthetic-generation capability. Access is available through a UI, APIs, and MCP-enabled workflows.
Market context
The synthetic and test-data management market has grown substantially as organisations scale AI-assisted development and face tighter data-privacy regulation. Competitors include incumbents such as IBM (Optim), Broadcom, and specialist vendors including Informatica and K2view, as well as a crop of AI-native startups positioning around privacy-preserving data generation. The sector sits at the intersection of DevOps tooling, data governance, and AI compliance, making it relevant to teams navigating both GDPR and emerging AI-Act obligations around training and testing data.
Perforce's decision to embed a bring-your-own-LLM architecture is a direct response to enterprise reluctance to send sensitive schema information to third-party model providers, a concern that has slowed adoption of cloud-hosted AI tooling in regulated industries including financial services and healthcare. Whether the natural-language interface meaningfully reduces configuration burden over legacy tools, beyond the company's own survey findings, remains to be tested by independent benchmarks. Perforce did not publish third-party validation of generation quality or referential-integrity fidelity in the launch materials.
Delphix Synthetic Data is available now as part of the Delphix DevOps Data Platform. Perforce did not disclose pricing or name any early-access customers in the release.